The present invention relates to an integrated database system having as a purpose to perform data processing utilizing a plurality of different databases.
A requirement to share or to commonly use databases is widely encountered within an organization, among organizations or among individuals. Further, because of the prevalence of wide-spread networks, regions sharing a group of databases are also extended. In such an environment, various attempts have already been made to realize sharing and common use of databases or practical commercialization within a limited area.
In order to enable sharing and common use of the group of databases not only within the aforementioned limited area, but also in a wider area, it is necessary to realize a multi-database system suitable for an environment where a large number of database systems are connected via a wide-spread high-speed computer network. Moreover, the cost of constructing such multi-database system must be as low as possible, and in addition, the multi-database system must be able to change freely after being constructed, and must be flexible in terms of necessary maintenance.
Herein, the "multi-database system" is a general term for a system which widely realizes data sharing within a distributed system or in a network system. The environment constructed by the multi-database system will be called "multi-database environment."
However, when a database is shared and commonly used, the following problems arise in general:
1. Data included in different databases may have different meanings even if the data is represented by the same representation form;
2. Vice versa to the above-mentioned problem 1, data included in different databases may be represented by different styles even if the data have the same meaning;
3. Although establishing a conventional manner of combining data is possible among different databases, the representation forms with regard to combining attributes may nonetheless be different. Herein, the term "conventional manner of combining data" means combining processing realized by pattern matching that has been adopted by the conventional database system, and the combining attributes mean attributes of interest at the time of combining);
4. Databases subjected to integration may lack necessary data for the combining process; and
5. Even if different types of plural databases are combined, new data cannot be generated therefrom.
As a partial solution to the above-described problems 1 to 4, the following processes have conventionally been executed:
1) To transform or absorb the difference in meanings of the data, difference in the representation forms, difference in the combining attributes and the like, codes are incorporated in an application program; and
2) To compensate for lacking data, the lacking data is incorporated in the application program as data or as a constant value, alternatively a database which compensates for the lacking data is newly added.
However, as a result of the above solution, the following problems are further raised:
6. The aforementioned application program is generated independently each time a command related to sharing and common use of the database is sent; thus, responding to the command takes time;
7. When the aforementioned command related to sharing and common use of the database is changed, for most cases, it is difficult or impossible to reuse the aforementioned application program.